- The CityGML encoding standard is a rich and valuable road map for building 3D urban geo-databases – kudos to all involved.
- LiDAR and laser scanning are the most appropriate technology for building out these databases.
- We can choose to take advantage of this or re-invent the wheel.
In preparing for my presentation for the 3D Fusion Summit I have had a chance to look a lot more closely at the CityGML standard. First of all, since I know first hand how difficult it is to produce something like this, I want to recognize the work of all involved, including the 3DIM Working Group, the OGC and the Special Interest Group 3D of the Initiative Geodata Infrastructure North-Rhine Westphalia. The latter is a group in Germany that is actually responsible for delivering the standard, and who, along with other groups in Germany, have been the leaders of this effort worldwide.
In addition, the 3DIM working group includes Autodesk, Bentley and the Ordnance Survey to name a few other groups. Bottom line – a number of major players are involved with this effort on a worldwide basis. Time to pay attention.
I think it is also important to note that CityGML is described as an encoding standard. If you take the time to read this document you will realize that this is a perfect description. The standard lays out the details of an XML schema for representing most of the elements of an urban landscape including buildings, city furniture, vegetation, water bodies, etc. If the industry buys in, there can be an elegant exchange of information among all stakeholders.
CityGML also specifies 5 levels of detail (LOD), another important data model organizational concept. These range from a 2.5 D digital terrain model to specifying doors and windows. The standard includes a table of dimensional accuracy associated with each LOD. It is noted that this is just preliminary and that more work needs to be done in this area. I agree.
LiDAR and laser scanning have the potential to be the catalyst for a 5 to 10 year effort to build out these 3D CityGML geo-databases. This is the only realistic approach to acquiring this data in a cost effective manner.
CityGML provides the road map. GIS never had this. We don’t have to reinvent the wheel, unless we want to.